JP2019507566A - Method for monitoring electromechanical actuator system - Google Patents

Abstract

The present invention relates to a method for monitoring an electromechanical actuator system by using a Kalman filter to estimate a voltage drop in a power supply to a motor that is linked to an inverter fault, and by taking into account the estimated voltage drop. It relates to a method comprising: at least estimating an electromagnetic torque coefficient (300); and calculating (400) a motor electromagnetic torque from the motor electromagnetic torque coefficient.
[Selection] Figure 2

Description

  The present invention relates to a method for monitoring an electromechanical actuator system.

  In an aircraft, various movable elements such as an aileron, flight control surface and even a reverse thrust device cover are associated with the actuator so that it can move between a neutral position and an operating position. For this purpose, each actuator is driven by a motor so as to translate between the two abutments. As an example, for a flap, the first abutment is associated with the neutral position of the movable element and the second abutment is associated with the operating position of the movable element. Under such circumstances, the motor drives the actuator when powered, which in turn moves the associated movable element.

  In order to detect any failure or aging of the drive mechanism of the movable element, it is appropriate to monitor the motor and actuator, which is usually done by sensors associated with the motor and / or actuator.

  Nonetheless, sensors that are usually placed around motors and / or actuators do not necessarily make it possible to obtain all of the desired measurements in order to monitor the various electrical and mechanical parameters of the drive mechanism. is not. For example, such a sensor does not make it possible to measure the electromagnetic torque of the motor, although it would be possible to deduce an advantageous parameter in the monitoring of the drive mechanism from the electromagnetic torque of the motor.

  It has been proposed to measure electromagnetic torque by using external measuring equipment when the aircraft is on the ground.

  Nevertheless, this requires the aircraft to be on the ground regularly for a sufficient length of time so that measurements can be taken and updated regularly to account for wear in the motor and actuator.

  Proposals have been made to integrate additional sensors in the aircraft to detect electromagnetic torque.

  Nevertheless, this inevitably leads to increased weight and bulk as well as cost, and these increases are particularly undesirable in the field of aircraft manufacturing.

  The object of the present invention is to propose an electromechanical actuator system monitoring method which avoids the above-mentioned drawbacks.

To this end, the present invention provides a method for monitoring an electromechanical actuator system comprising at least one inverter, a motor powered by the inverter, and an actuator driven by the motor.
A motor linked to an inverter fault using a Kalman filter taking into account operating data including at least one current delivered by the inverter to the motor, at least one control voltage of the inverter and at least one rotational speed of the motor exit shaft Estimating a voltage drop in the power supply to
Consider operating voltage data including estimated voltage drop and current sent by inverter to motor, inverter control voltage, derivative of current sent by inverter to motor and rotational speed of motor exit shaft And at least estimating an electromagnetic torque coefficient of the motor;
Calculating the electromagnetic torque of the motor from the operation data constituted by the electromagnetic torque coefficient of the motor and the current sent by the inverter to the motor;
A method comprising:

  The method of the present invention makes it possible to estimate the electromagnetic torque by direct calculation, thus avoiding carrying a heavy and expensive specific sensor with an electromechanical actuator system for this magnitude measurement. To. Thus, from this electromagnetic torque, arbitrarily various additional mechanical parameters, such as the efficiency of the motor / actuator assembly, and the viscous friction coefficient and dry friction torque, both representing the mechanical friction in the motor / actuator assembly, are deduced. Is possible. This guess of electromagnetic torque advantageously takes into account uncertainty about disturbances due to faults in the inverter that cause switching and conduction losses as well as voltage drops produced by fault periods. Thus, it has been found that the estimation of electromagnetic torque is relatively accurate.

  Similarly, the electromagnetic torque is calculated directly during operation, and therefore no specific operation is required that needs to be taken off of the device carrying the electromechanical actuator system for implementation.

  Throughout this application, the term “motion data” is obtained during a specific operation of an electromechanical actuator system intentionally performed to obtain data, and the electromechanical actuator system is intentionally shut down for that purpose. Is used to refer to data acquired while the electromechanical actuator system is in operation and installed in its natural state. Thus, when the electromechanical actuator system is mounted on an aircraft, the method of the present invention can be used as operational data during flight (eg, to extract data) rather than during maintenance work performed while the aircraft is on the ground. Use data measured during take-off, landing, turning, etc. when these operations are not intentionally performed to mount the actuator for the purpose.

  In certain embodiments, the method adds an estimate of at least one mechanical parameter of the electromechanical actuator system from electromagnetic torque, operational data including at least the rotational speed of the motor exit shaft, and aerodynamic forces experienced by the actuator. Includes steps.

  In certain embodiments, the mechanical parameter is the coefficient of viscous friction and / or the dynamic drive friction torque and / or the efficiency of the assembly including the motor and actuator.

  In certain embodiments, the method includes the additional steps of creating a database and filling at least an estimate of the electromagnetic torque of the motor in the database.

  In certain embodiments, the database similarly includes at least motor resistance and / or motor stator inductance and / or viscous friction coefficient of the motor / actuator assembly and / or dynamic dry friction torque of the motor / actuator assembly and / or Alternatively, electrical and mechanical parameters other than the motor electromagnetic torque, including the direct efficiency of the motor / actuator assembly and / or the indirect efficiency of the motor / actuator assembly are also entered.

  Other features and advantages of the present invention will become apparent from the following description of specific embodiments of the invention.

  The invention can be better understood in light of the following description of certain non-limiting embodiments of the invention. Reference is made to the accompanying drawings.

FIG. 6 is a schematic diagram of an electromechanical actuator system implementing the method in a particular embodiment of the invention. 2 is a schematic diagram illustrating various steps of a method performed by the system schematically illustrated in FIG. FIG. 2 is a circuit diagram modeling the electrical part of the system shown in FIG. 1 within one frame, respectively in relation to the D axis and the Q axis. FIG. 2 is a circuit diagram modeling the electrical part of the system shown in FIG. 1 within one frame, respectively in relation to the D axis and the Q axis.

  1 and 2, the monitoring system in a particular embodiment of the present invention is applied in this example to an electromechanical actuator system 1 for activating an aircraft aileron A. Of course, this application is not limiting, and the method of the present invention can be applied to several other electromechanical actuators, such as electromechanical actuator systems that are coupled to aircraft reverse thruster covers, flaps or flight control surfaces, etc. It can be implemented in the system.

  In this embodiment, the electromechanical actuator system 1 includes an electrical portion 10 that includes an inverter 11 and a motor 12 that is powered by the inverter 11. In this embodiment, the motor 12 is a permanent magnet synchronous motor. The electromechanical actuator system 1 similarly includes a motor 12 that includes a motor 12 with an actuator 21 coupled first to the outlet shaft of the motor 12 and secondly to the auxiliary wing A so that the auxiliary wing A can be moved. A portion 20 is also included. In this embodiment, the actuator 21 is of a linear type, and as an example, the actuator includes a ball screw type or roller screw type jack. In a variant, the actuator can be of the rotary type. Under such a situation, the motor 12 drives the actuator 21 at the time when electric power is supplied, and this actuator moves the auxiliary wing A that is associated therewith.

  Thus, the method of the present invention serves to monitor the electromechanical actuator system 1 in the manner detailed below.

  Since the motor 12 receives the three-phase AC power supplied by the inverter 11, the electrical portion 10 is modeled in one frame (straight axis and horizontal axis) as shown in FIGS. 3a and 3b. Thus, the various electrical magnitudes calculated or measured in the method are projected onto the two principal axes of the two-phase model, namely the D axis (straight axis) and the Q axis (horizontal axis). In the following description, index I means one magnitude projection on the D axis, and index Q means one magnitude projection on the q axis.

For the purpose of estimating the voltage drop in the power supply of the motor 12 associated with a failure of the inverter 11, the first step 100 of the method comprises:
Continuous
,
Discrete
,
The construction of a linear, stationary and stochastic Kalman filter of the type

  Such Kalman models are well known to those skilled in the art and are therefore not described in detail herein. For further information, see P.I. S. Maybek's book “Stochastic Models, Estimation and Control”, Volume 141-1, Mathematical in Science and Engineering can be instructed.

To apply the Kalman estimator, the following equation is expressed for the synchronous motor 12 in the frame:
_Where Y _D and V _Q are power supply voltages that are controlled to be _delivered by the inverter 11, called control voltages (measured in the current regulator loop of the inverter 11 in this embodiment), and V _Dinverter And V _Qinverter are inverter voltage drops due to inverter defects, and ΔV _D and ΔV _Q are uncertainties regarding inverter voltage drops due to various electrical parameter uncertainties in electrical portion 10. I _D and i _Q are currents transmitted to the motor 12 by the inverter 11 (measured in the motor 12 in this embodiment), R is the resistance of the motor 12, and L _D and L _Q Is the inductance of the stator phase of the motor 12, p is the number of pole pairs of the motor 12, ω is the rotational speed of the outlet shaft of the motor 12, and Ke is the electromagnetic torque coefficient of the motor 12. A.

Data i _D , i _Q , V _D and V _Q and ω can be either operational data measured by a sensor, or a memory (eg, for later use in real time or measured / recovered and subsequently used by the method of the present invention. Stored in the computer's memory) includes operational data retrieved from commands transmitted by the computer to the inverter 11 while the aircraft is in flight.

Prior to establishing the Kalman model for the purpose of estimating the voltage drops V _Dinverter and V _Qinverter due to inverter defects, the following three assumptions are considered:

In the first assumption, voltage drops V _Dinverter and V _Qinverter due to inverter defects were associated with random variables W _D (t) and W _Q (t) white noise types with known spectral power density without bias. It is an integer type variable. Ie

In the second assumption, the following variable changes are considered to allow decoupling.

  In the first assumption, all Kalman convergence conditions are satisfied. For more detailed information, please refer to P.C. S. Maybek's book, "Stochastic Models, Estimation and Control", Volume 141-1, Mathematical in Science and Engineering can be directed.

Thus, the final model for estimating the voltage drops V _Dinverter and V _Qinverter due to the defect of the inverter 11 is provided by the following state representation.

Thus, for the two axes d and q,
And for the output matrix,
Thus, the state expression in the general forms (1) and (2) can be found.

  The method thus makes it possible to establish two Kalman models for the electromechanical actuator system 1 along the axis d and the axis q that are appropriate to process independently.

For these two Kalman models, the constants R, L _Q and L _D and Ke are theoretical values. Uncertainties regarding these values are considered external disturbances, which have already been taken into account in the two Kalman models. As an example, it is possible to rely on manufacturer data in imposing these values. Similarly, p is known from, for example, manufacturer data.

  Thus, once the Kalman models are established, the method includes a second step 200 of determining the states x (t) of these models by a Kalman filter.

  For this purpose, the model defined above is used in discrete form in this embodiment.

  Applying the Kalman model to a discrete form and applying the Kalman algorithm to the resulting discrete model is well known to those skilled in the art and is therefore not described in detail herein. For further details on making the two Kalman models into discrete form and the Kalman algorithm's recurrence equations, see, for example, p. S. Mayback's book, “Stochastic Models, Estimation and Control”, Volume 141-1, Mathematical in Science and Engineering can be directed.

Thus, the known Kalman algorithm applied to the two discrete Kalman models obtains an guess for x (t), so that the motion data i _D , i _Q on the two axes d and q, , VD, VQ, and ω from the voltage drop V _Dinverter and V _Qinverter of the inverter 11 can be obtained. Once the voltage drop of the inverter 11 is estimated, the method performs a third step 300 for estimating various electrical parameters associated with the electromechanical actuator system 1 including at least the estimated motor torque constant Ke _est. Have Preferably, the method serves to estimate other electrical parameters, namely the resistance R _{est of} the motor 12 and the inductance L _{est of the} motor 12, and in the first step the values R, L _D and L _Q It is understood that is set to a nominal theoretical value selected based on manufacturer data.

To this end, it should be recalled that the electrical equation (3) presented above can be expressed using actual parameters in the following form.

It is assumed that L _Qreal = L _Dreal = L _real .

Based on this assumption, from the system of equations (4), the measured values measured nth and used in the method (the current sent by the inverter 11 to the motor 12, the control voltage, and the rotational speed of the outlet shaft of the motor 12). The following linear equation system (5) corresponding to
This equation can also be expressed in the form X _n = h _n θ _nreal , where
It is.

Nevertheless, as already mentioned, it can be seen that the measured values that the electromechanical actuator system 1 measures and uses are usually noisy. In order to make the system of equations (5) more realistic by taking these uncertainties about the measurements into account, a vector ν _n corresponding to unbiased white noise is introduced into the system of equations. Thus, the following relationship is obtained.
Y _n = h _n θ _nest −V _n
In the formula

The data V _DInverter and V _QInverter have been estimated in the preceding step 200, the data p is known (from manufacturer data), and the data i _D , i _Q , V _D , V _Q and ω are measured. All that remains is to determine the parameter vector θ _nest .

In this embodiment, this determination is performed by probabilistic calculation. More precisely, this determination is performed by a recursive least squares algorithm that makes it possible to determine θ _nest while minimizing the criterion ε _n = (Y _n −X _n ) ^T (Y _n −X _n ). The

  Such algorithms are well known to those skilled in the art and are therefore not described in detail herein. For further details, see, for example, P.I. S. Mayback's book, “Stochastic Models, Estimation and Control”, Volume 141-1, Mathematical in Science and Engineering can be directed.

Thus, this makes it possible to estimate the motor electromagnetic torque coefficient Ke _est , the motor resistance R _est, and the motor inductance L _est . Thereby, the health status of the electromechanical actuator system 1 is monitored on the basis of known aging relationships of the electromechanical actuator system 1 and based on these estimated parameters and / or the aircraft is then brought to the ground. It is possible to arbitrarily perform preventive maintenance work when parked.

Once the motor electromagnetic torque coefficient Ke _est has been estimated, the method can be expressed as
A fourth step 400 of calculating the electromagnetic torque C _{elec of the} motor from In the equation, Δi _Q is an error in the current measurement value that is bounded.

Preferably, the method comprises a plurality of mechanical parameters associated with the electromechanical actuator system 1, namely the viscous friction coefficient Coef _{f-visc of} the mechanical part 20, the dynamic dry friction torque C _{f-dry of the} mechanical part 20. And the direct efficiency ρ _direct of the mechanical part 20 (ie connected to the motor as opposed to the _indirect efficiency ρ _indirect corresponding to the efficiency when the load connected to the motor is associated with the movement of the motor outlet shaft) Is the efficiency when the applied load counteracts the movement of the motor outlet shaft, the relationship between direct efficiency and indirect efficiency is
A fifth step 500 of estimating

For that purpose, the basic principle of dynamics is applied to the outlet shaft of the motor 12,
Where you get
Where F is the aerodynamic force experienced by the actuator 21 measured using a sensor carried by the actuator 21 and J _mom is the moment of inertia of the system as seen from the shaft of the motor 12.

It can be expressed as follows.

This gives the following linear equation corresponding to the nth measurement value measured and used in the method.
This equation can also be expressed in the form X _n = h _n θ _n , where
It is.

Nevertheless, as already mentioned, it can be seen that the measured values that the electromechanical actuator system 1 measures and uses are usually noisy. Because of the current measurement error Δi _Q , there remains uncertainty regarding the motor electromagnetic torque _Celec . In order to make equation (6) more realistic, equation (6) is modified to take this uncertainty into account. In this example, the uncertainty _Celec regarding the electromagnetic torque of the motor is expressed by the white noise ν _{n at} the center of the known spectral power density.

Thus, the following equation is obtained:
Y _n = h _n θ _nest + ν _n , where

The angular acceleration dω / dt in this example is calculated from the angular velocity ω, but in one variant it can be measured at the motor 12. Similarly, knowing that the data Ke has been estimated in the first step 300, the data J _mom and pitch _screw are known, and the data ω, i _Q , F are measured motion data. Thus, all that remains is to determine the parameter vector θn _est .

In this embodiment, this determination is performed by probabilistic calculation. More precisely, this determination is performed by a recursive least squares algorithm that makes it possible to determine θ _n while minimizing the criterion ε _n = (Y _n −X _n ) ^T (Y _n −X _n ). The

  Moreover, it should be recalled that the algorithm needs to take into account the fact that the preceding equation is applicable only when the rotational speed of the motor 12 is not zero. Under such circumstances, if the rotational speed of the motor 12 in absolute value is higher than a certain threshold defined as arbitrarily close to zero, the parameter estimated in the algorithm is “frozen” to the closest value. Is done. Preferably, the pressure is adjusted to be as close to zero as possible as the accuracy of the Kalman filter increases.

Assuming that P is an autocorrelation matrix of estimation error ε _n and R _v is a covariance matrix of noise ν _n , the algorithm is expressed as follows.
If ω (n) <threshold or if threshold <ω (n),
K _n = K _n-1
θ _nest = θ _n-lest
P _n = P _n-1
Otherwise,
When n = 1 (initial condition),
Otherwise,

Thus, this makes it possible to estimate the viscous friction coefficient Coef _f-visc , dynamic dry friction torque C _f-dry , and direct efficiency ρ _direct . This makes it possible to monitor the health state of the electromechanical actuator system 1 from the known aging relationship for the electromechanical actuator system 1 and for the estimated parameters. As an example, it is thus possible to estimate the aging of the electromechanical actuator system 1 and / or optionally perform preventive maintenance work while the aircraft is parked on the ground.

  Thus, the method described above makes it possible to estimate various electrical and mechanical parameters suitable for monitoring the health state of the electromechanical actuator system 1. As described above, the method is performed during the actual flight of the aircraft, not during operations dedicated to estimating these parameters or while the aircraft is not in service.

  The method can thus be carried out directly during the operation of the aircraft that is required to operate the electromechanical actuator system 1 and even after this operation has been performed, where it is required to carry out the method. Various measurements are pre-recorded during operation so that they can be used retrospectively by the method.

  Thus, the data obtained by the method of the present invention makes it possible to implement various strategies for monitoring the electromechanical actuator system 1 and taking action on the electromechanical actuator system 1. For example, the method may include the additional step of forming a database composed of various values for electrical and mechanical parameters estimated over time for different flights of the aircraft. This database thus makes it possible to determine how the various electrical and mechanical parameters of the system vary over time. The database can also serve to facilitate monitoring of other electromechanical actuator systems.

  Naturally, the invention is not limited to the embodiments described, but alternative embodiments may be provided without departing from the scope of the invention as defined by the claims.

Claims (5)

In a method of monitoring an electromechanical actuator system comprising at least one inverter (11), a motor (12) powered by the inverter, and an actuator (21) driven by the motor,
The method
Using a Kalman filter in consideration of operation data including at least one current that the inverter sends to the motor, at least one control voltage of the inverter, and at least one rotational speed of the outlet shaft of the motor Estimating (100, 200) a voltage drop in the power supply to the motor associated with a fault in the inverter;
The estimated voltage drop, the current that the inverter sends to the motor, the control voltage of the inverter, the derivative of the current that the inverter sends to the motor, and the motor Estimating (300) at least an electromagnetic torque coefficient of the motor by considering the operational data including the rotational speed of the outlet shaft;
Calculating (400) the electromagnetic torque of the motor from operational data comprising the electromagnetic torque coefficient of the motor and the current sent by the inverter to the motor.   2. The method according to claim 1, wherein simultaneously with estimating the electromagnetic torque coefficient of the motor, other electrical parameters of the electromechanical actuator system (1) are estimated as well.   The method of claim 2, wherein the other estimated electrical parameters are the resistance of the motor and the inductance of the motor.   The method is based on the electromagnetic torque, the operation data including at least the rotational speed of the outlet shaft of the motor (12), and the aerodynamic force received by the actuator (21). The method according to any one of the preceding claims, comprising an additional step (500) of estimating at least one mechanical parameter.   The method of claim 4, wherein the mechanical parameter is a viscous friction coefficient and / or a dynamic drive friction torque and / or an efficiency of an assembly including the motor and the actuator.